mmmmm...
This reminds me of a rather heated discussion that I started about 6 months ago, hey Tarq57? Although my agenda at the time was coming at this subject from a slightly different angle, and in the end I stirred the hornets nest intentionally to get people talking, I came to the conclusion that EFPS is not a mature product at this time.
It kind of falls into the same basket as voice recognition, sounds really good, but unless its smart enough to work out the difference between "number 1 at the holding point" and "the red piper holding short of RWXX" its not ready. The conclusion I came to is its fine for some aspects of ATC, especially en-route, but throw in VFR, military etc, and even the most mature of the systems (and I looked closely at a lot of them) are still left wanting.
One day though, I have no doubt that it will be the standard, or more likely voice recognition and EFPS will integrate seamlessly, I just cringe thinking that some controllers are basically working in a live lab that is feeding the industrial system engineers. Lets hope the worst doesn't happen due to an unforeseen piece of Swiss cheese...
The only part of the argument that I feel should be closed off is system reliability, if its designed properly, with the right engineering standards applied, and with the correct system maintenance planned for, it will be rock solid.